p16 protein expression and human papillomavirus status as prognostic biomarkers of nonoropharyngeal head and neck squamous cell carcinoma.

Published Ahead of Print on October 6, 2014 as 10.1200/JCO.2013.54.5228 The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2013.54.5228

JOURNAL OF CLINICAL ONCOLOGY

O R I G I N A L

R E P O R T

p16 Protein Expression and Human Papillomavirus Status As Prognostic Biomarkers of Nonoropharyngeal Head and Neck Squamous Cell Carcinoma Christine H. Chung, Qiang Zhang, Christina S. Kong, Jonathan Harris, Elana J. Fertig, Paul M. Harari, Dian Wang, Kevin P. Redmond, George Shenouda, Andy Trotti, David Raben, Maura L. Gillison, Richard C. Jordan, and Quynh-Thu Le Author affiliations appear at the end of this article. Published online ahead of print at www.jco.org on September 29, 2014. Supported by Radiation Therapy Oncology Group Grant No. U10 CA21661 and National Cancer Institute Community Clinical Oncology Program Grant No. U10 CA37422. Presented orally at the 49th Annual Meeting of the American Society of Clinical Oncology, Chicago, IL, May 31-June 4, 2013. Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article. Clinical trial information: NCT00047008, NCT00265941, NCT00084318. Corresponding author: Christine H. Chung, MD, Johns Hopkins University, 1650 Orleans St, CRB-1 Room 344, Baltimore, MD 21287-0013; e-mail: [email protected]. © 2014 by American Society of Clinical Oncology 0732-183X/14/3209-1/$20.00 DOI: 10.1200/JCO.2013.54.5228

A

B

S

T

R

A

C

T

Purpose Although p16 protein expression, a surrogate marker of oncogenic human papillomavirus (HPV) infection, is recognized as a prognostic marker in oropharyngeal squamous cell carcinoma (OPSCC), its prevalence and significance have not been well established in cancer of the oral cavity, hypopharynx, or larynx, collectively referred as non-OPSCC, where HPV infection is less common than in the oropharynx. Patients and Methods p16 expression and high-risk HPV status in non-OPSCCs from RTOG 0129, 0234, and 0522 studies were determined by immunohistochemistry (IHC) and in situ hybridization (ISH). Hazard ratios from Cox models were expressed as positive or negative, stratified by trial, and adjusted for clinical characteristics. Results p16 expression was positive in 14.1% (12 of 85), 24.2% (23 of 95), and 19.0% (27 of 142) and HPV ISH was positive in 6.5% (six of 93), 14.6% (15 of 103), and 6.9% (seven of 101) of non-OPSCCs from RTOG 0129, 0234, and 0522 studies, respectively. Hazard ratios for p16 expression were 0.63 (95% CI, 0.42 to 0.95; P ⫽ .03) and 0.56 (95% CI, 0.35 to 0.89; P ⫽ .01) for progression-free (PFS) and overall survival (OS), respectively. Comparing OPSCC and non-OPSCC, patients with p16-positive OPSCC have better PFS and OS than patients with p16-positive non-OPSCC, but patients with p16-negative OPSCC and non-OPSCC have similar outcomes. Conclusion Similar to results in patients with OPSCC, patients with p16-negative non-OPSCC have worse outcomes than patients with p16-positive non-OPSCC, and HPV may also have a role in outcome in a subset of non-OPSCC. However, further development of a p16 IHC scoring system in non-OPSCC and improvement of HPV detection methods are warranted before broad application in the clinical setting. J Clin Oncol 32. © 2014 by American Society of Clinical Oncology

INTRODUCTION

Head and neck squamous cell carcinoma (HNSCC) is a heterogeneous disease occurring in various sites within the head and neck region, including the oral cavity, oropharynx, hypopharynx, and larynx. The most common risk factors are tobacco and alcohol use and high-risk human papillomavirus (HPV) infection.1,2 Although the detection rates vary depending on assay selection and study populations, approximately 57% to 72% of oropharyngeal squamous cell carcinomas (OPSCCs) and 1.3% to 7% of non-OPSCCs, including cancers of the oral cavity, hypopharynx, and larynx, are HPV positive.3-8

HPV status in tumors can be determined by several assays, including HPV DNA detection by in situ hybridization (ISH) or polymerase chain reaction (PCR), HPV E6/E7 RNA expression detected by quantitative reverse transcriptase–PCR (qRT-PCR), and/or p16 protein expression by immunohistochemistry (IHC) staining as a surrogate marker of oncogenic HPV infection.5-7,9-11 Among these assays, detection of HPV E6/E7 RNA expression, which indicates active viral oncogene transcription in tumor cells, is considered to be a gold standard.9,10 However, because RNA isolation for qRT-PCR requires additional sample preparation steps and a larger amount of tumor cells compared with other © 2014 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at UCSF LIBRARY & CKM on April 6, 2015 from Copyright © 2014 American Society of Clinical Oncology. All rights reserved. 169.230.243.252

Copyright 2014 by American Society of Clinical Oncology

1

Chung et al

assays, the most widely used assays are HPV ISH and p16 IHC. When the results of various assays are compared, the concordance rate between HPV ISH and p16 IHC is approximately 90% in OPSCC, where HPV infection is frequent.5 However, in oral cavity squamous cell carcinoma (SCC), where infection is relatively less common than in the oropharynx, sensitivity of p16 IHC compared with high-risk HPV E6/7 RNA expression is 79%, specificity is 93%, positive predictive value (PPV) is 41%, and negative predictive value is 99%, indicating that p16 IHC is a poor surrogate biomarker of HPV infection in non-OPSCC sites.7 It is also well established that patients with HPV-positive/p16positive OPSCC have a more favorable prognosis compared with those with HPV-negative/p16-negative OPSCC.5,6,11,12 However, the prognostic significance of p16 expression in non-OPSCC with or without evidence of HPV infection has not been clearly delineated. The p16 protein is an important tumor suppressor and cell-cycle regulator.13 In HPV-positive tumors, the viral protein E7 binds to retinoblastoma susceptibility protein (Rb) through cullin 2 ubiquitin ligase complex and rapidly degrades Rb by ubiquitination.14,15 Loss of Rb results in upregulation of p16 protein expression by a feedback interaction.16,17 However, increased p16 protein expression is not specific to Rb loss caused by E7 oncoprotein, because other molecular events associated with loss of Rb function, such as RB1 inactivating mutation, or deletion or chromosomal loss, can result in the same phenotype. In this study, we evaluated p16 protein expression by IHC and HPV status by HPV ISH as potential prognostic biomarkers in nonOPSCC tumors, where HPV infection is less common than in the oropharynx, in patients enrolled onto three prospective Radiation Therapy Oncology Group (RTOG) clinical trials.

PATIENTS AND METHODS Protocol RTOG 0129 was a phase III trial evaluating standard fractionation radiotherapy with concurrent cisplatin versus accelerated fractionation by concomitant boost radiotherapy with concurrent cisplatin for patients with locally advanced HNSCC (N ⫽ 743).5 RTOG 0234 was a phase II randomized trial testing whether radiation therapy with concurrent cetuximab and either cisplatin or docetaxel improved disease-free survival over a historical cohort of patients treated with radiation therapy and concurrent cisplatin (RTOG 9501) for patients at high risk for recurrence after surgical resection of advanced HNSCC (N ⫽ 238).18 RTOG 0522 was a phase III trial testing the addition of cetuximab to radiation therapy with concurrent cisplatin for patients with locally advanced HNSCC (N ⫽ 940).19

Laboratory Studies IHC was performed to determine p16 expression using a p16 mouse monoclonal antibody (predilute, mtm-CINtech, E6H4) as previously described.5 p16 was considered to be positive when defined as strong and diffuse nuclear and cytoplasmic staining in ⱖ 70% of the tumor cells, which is the same scoring criteria used in the study by Ang et al.5 High-risk HPV status was determined by ISH using a cocktail probe that detects HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 66 (GenPoint HPV probe cocktail; Dako, Carpinteria, CA).5 HPV ISH was interpreted as positive when nuclear-specific staining was detected in the tumor cells. 2

© 2014 by American Society of Clinical Oncology

Statistical Analysis The end points, progression-free (PFS) and overall survival (OS), were as defined in the RTOG 0522 protocol. Nonachievement of PFS was defined as locoregional recurrence or progression, distant metastasis, or death resulting from any cause. Nonachievement of OS was defined as death resulting from any cause. Rates were estimated by the Kaplan-Meier method20 and compared with a two-sided stratified (by study) log-rank test.21 Univariable and multivariable hazard ratios (HRs) were estimated using the Cox proportional hazards model,22 stratified by study. The ␾ coefficient was used to measure the association between HPV and p16 status.

RESULTS

Expression of p16 and High-Risk HPV ISH Status Determination A total of 683 eligible patients with non-OPSCC tumors, including primary sites in the oral cavity, hypopharynx, and larynx, were identified among the 1,921 patients enrolled onto RTOG 0129, 0234, and 0522 studies. Tumors from 356 (52.1%) of 683 patients with non-OPSCC were tested for p16 expression, which could be determined in 90.4% (322 of 356) of the tumors. The results are summarized in Table 1. Characteristics and outcomes were similar between patients with known and unknown p16 status in each trial (Appendix Table A1, online only). Overall, 19.3% (62 of 322) were p16 positive. The rates of p16 positivity were 14.1%, 24.2%, and 19.0% for RTOG 0129, 0234, and 0522, respectively. Cancer of the oral cavity had the highest rate of p16 positivity (21 [26.3%] of 80), followed by the larynx (31 [17.1%] of 181) and hypopharynx (10 [16.4%] of 61). Patient characteristics by p16 expression status are summarized in Appendix Table A2 (online only). A total of 311 (45.5%) of 683 tumors had tumor material available for high-risk HPV testing by ISH (Table 1). Of these, HPV status could be determined in 95.5% (297 of 311). Overall, 9.4% (28 of 297) were HPV ISH positive. The HPV ISH–positive rates were 6.5%, 14.6%, and 6.9% for RTOG 0129, 0234, and 0522, respectively. Cancer of the oral cavity had the highest rate of HPV ISH positivity (13 [14.6%] of 89), followed by the larynx (12 [7.9%] of 151) and hypopharynx (three [5.3%] of 57). Patient characteristics by high-risk HPV ISH status are summarized in Appendix Table A3 (online only). When the two tests were compared, there was moderate correlation between p16 and HPV status (␾ coefficient, 0.46; 95% CI, 0.34 to 0.58; Table 2). The tests were concordant more often when the primary site was the hypopharynx (␾ coefficient, 0.54; 95% CI, 0.28 to 0.81) or larynx (␾ coefficient, 0.52; 95% CI, 0.35 to 0.68) compared with the oral cavity (␾ coefficient, 0.35; 95% CI, 0.13 to 0.57). Survival Outcomes Based on p16 and HPV ISH Status We examined survival outcome differences based on p16 status in patients with non-OPSCC tumors. Univariable analysis showed that patients with p16-positive tumors had better PFS and OS than patients with p16-negative tumors (Figs 1A and 1B). For PFS, the HR estimate was 0.65 (95% CI, 0.44 to 0.98), reflecting a 35% reduction in PFS nonachievement for patients with p16-positive tumors (P ⫽ .04). For OS, the HR was 0.57 (95% CI, 0.36 to 0.90), reflecting a 43% reduction in the death rate for patients with p16-positive tumors (P ⫽ .01). Furthermore, the effects of p16 expression in PFS and OS were examined after adjustment for known prognostic factors, including age, sex, T stage, and N stage (Table 3). The adjusted HRs were 0.63 JOURNAL OF CLINICAL ONCOLOGY


p16 Protein Expression and HPV Status in Nonoropharyngeal HNSCC

Table 1. Summary of p16 Expression and High-Risk HPV Status RTOG 0129 (n ⫽ 288) Status HPV tested Tested, value known Tested, value unknown Not tested HPV status Negative Positive HPV status, oral cavity Negative Positive HPV status, hypopharynx Negative Positive HPV status, larynx Negative Positive p16 tested Tested, value known Tested, value unknown Not tested p16 status Negative Positive p16 status, oral cavity Negative Positive p16 status, hypopharynx Negative Positive p16 status, larynx Negative Positive

RTOG 0234 (n ⫽ 129)

RTOG 0522 (n ⫽ 266)

Total (N ⫽ 683)

No.

%

No.

%

No.

%

No.

%

93 10 185

32.3 3.5 64.2

103 1 25

79.8 0.8 19.4

101 3 162

38.0 1.1 60.9

297 14 372

43.5 2.0 54.5

93.5 6.5

88 15

85.4 14.6

94 7

93.1 6.9

269 28

85.7 14.3

64 11

100.0 0.0

6 1

55 4

93.2 6.8

85 18 185

n ⫽ 93 87 6

n ⫽ 103

n ⫽ 14 12 2

18 3

85.7 14.3

29.5 6.3 64.2

95 4 30

85.9 14.1

72 23

66.7 33.3

51 17

100.0 0.0

6 1

84.9 15.1

15 5

n ⫽ 57 54 3

66 5

93.0 7.0

139 12

92.1 7.9

73.6 3.1 23.3

142 12 112

53.4 4.5 42.1

322 34 327

47.1 5.0 47.9

75.8 24.2

115 27

81.0 19.0

260 62

n ⫽ 71

n ⫽ 95

94.7 5.3 n ⫽ 151

n ⫽ 142

n ⫽ 322

n ⫽ 0ⴱ

n ⫽ 68

80.7 19.3 n ⫽ 80

75.0 25.0

59 21

n⫽7

n ⫽ 53 45 8

85.4 14.6

93.3 6.7

n ⫽ 21

n ⫽ 20 20 0

n ⫽ 30 28 2

n ⫽ 12 8 4

76 13

85.7 14.3

n ⫽ 85 73 12

90.6 9.4 n ⫽ 89

85.3 14.7 n⫽7

n ⫽ 59

n ⫽ 297

n ⫽ 0ⴱ

n ⫽ 75

n ⫽ 20 20 0

n ⫽ 101

n ⫽ 34 85.7 14.3

25 9

75.0 25.0

90 18

n ⫽ 20

73.8 26.3 n ⫽ 61

73.5 26.5

51 10

83.3 16.7

150 31

n ⫽ 108

83.6 16.4 n ⫽ 181 82.9 17.1

Abbreviations: HPV, human papillomavirus; RTOG, Radiation Therapy Oncology Group. ⴱ Patients with oral cavity tumors were not eligible for RTOG 0522.

(95% CI, 0.42 to 0.95; P ⫽ .03) and 0.56 (95% CI, 0.35 to 0.89; P ⫽ .01) for PFS and OS, respectively. We also examined survival outcome differences based on HPV ISH in these patients. Univariable analysis showed that patients with HPV ISH–positive tumors did not have significantly better PFS (HR, 0.77; 95% CI, 0.44 to 1.33; P ⫽ .35) or OS (HR, 0.64; 95% CI, 0.34 to 1.21; P ⫽ .17) than patients with HPV ISH–negative tumors (Figs 1C and 1D). In addition, PFS and OS in p16-positive/HPV-positive, p16-positive/HPV-negative, p16-negative/HPV-positive, and p16negative/HPV-negative groups were evaluated (Figs 1E and 1F); no significant differences were found (PFS, P ⫽ .23; OS, P ⫽ .21), probably because of the small sample sizes for the first three groups (n ⫽ 20, 33, and 7, respectively) compared with the p16-negative/HPVnegative group (n ⫽ 213). Survival Outcomes Based on p16 Status and Primary Site We also examined survival outcomes based on p16 status and three primary sites (oral cavity, hypopharynx, or larynx) separately. Although only PFS in the hypopharynx reached statistical significance because of the small sample size in each site, more favorable outcomes www.jco.org

in p16-positive patients compared with p16-negative patients were suggested (Appendix Figs A1 and A2, online only). For the oral cavity, the HRs for PFS and OS were 0.8 (95% CI, 0.42 to 1.55; P ⫽ .52) and 0.70 (95% CI, 0.33 to 1.47; P ⫽ .34), respectively. For the larynx, the HRs for PFS and OS were 0.61 (95% CI, 0.33 to 1.13; P ⫽ .11) and 0.54 (95% CI, 0.27 to 1.10; P ⫽ .08), respectively. For the hypopharynx, the HRs for PFS and OS were 0.33 (95% CI, 0.12 to 0.96; P ⫽ .03) and 0.33 (95% CI, 0.10 to 1.09; P ⫽ .06), respectively, favoring p16-positive patients. We further examined survival outcomes based on p16 status and primary sites, comparing non-OPSCC with OPSCC to examine the survival difference in non-OPSCC in the context of OPSCC, where the prognostic power of p16 is well established. p16 expression data in OPSCC from the three RTOG trials have been previously published.5,18,19 With both PFS and OS, there was a significant interaction between p16 status and primary site (interaction P ⫽ .01 for both; Table 4; Fig 2). Patients with p16-positive OPSCC had half the risk of death when compared with patients with p16-positive non-OPSCC (OS: HR, 0.48; 95% CI, 0.30 to 0.78). However, patients with p16negative OPSCC and non-OPSCC had similar survival, even after © 2014 by American Society of Clinical Oncology


3

Chung et al

Table 2. Association Between p16 Expression and High-Risk HPV Status p16 Status Negative HPV Status All primary sites Negative Positive Total Correlation (␾ coefficient) 95% CI Oral cavity only Negative Positive Total Correlation (␾ coefficient) 95% CI Hypopharynx only Negative Positive Total Correlation (␾ coefficient) 95% CI Larynx only Negative Positive Total Correlation (␾ coefficient) 95% CI

Positive

Total

No.

%

No.

%

No.

%

213 7 220

78.0 2.6 80.6

33 20 53

12.1 7.3 19.4

246 27 273

90.1 9.9 100.0

16.5 10.1 26.6

66 13 79

83.5 16.5 100.0

11.1 5.6 16.7

51 3 54

94.4 5.6 100.0

10.0 6.4 16.4

129 11 140

92.1 7.9 100.0

0.46 0.34 to 0.58 53 5 58

67.1 6.3 73.4

13 8 21 0.35 0.13 to 0.57

45 0 45

83.3 0.0 83.3

6 3 9 0.54 0.28 to 0.81

115 2 117

82.1 1.4 83.6

14 9 23 0.52 0.35 to 0.68

Abbreviation: HPV, human papillomavirus.

adjustment for other prognostic variables (OS: HR, 0.97; 95% CI, 0.74 to 1.24). Although these data are intriguing, there is significant heterogeneity within the non-OPSCC patient population that is inherent to the complex anatomy of the head and neck region, with numerous anatomic subsites and varying clinical outcomes (Appendix Table A4, online only). Survival Outcomes Based on p16 Status and Smoking Status Seventy-five percent of patients with non-OPSCC with p16positive tumors and 84% of those with p16-negative tumors had ⬎ 10 pack-years of exposure. Median number of pack-years was 33.8 (interquartile range, 10 to 51) and 40 (interquartile range, 20 to 60) for patients with p16-positive and p16-negative non-OPSCC tumors, respectively; the difference was not statistically significant (P ⫽ .12; Appendix Table A2, online only). Adding pack-years to the multivariable models did not change HRs for p16 status, suggesting that survival outcome differences observed between the p16-positive and p16negative patients were not a function of smoking history (data not shown). DISCUSSION

Similar to results in patients with OPSCC, our data suggest that patients with p16-positive non-OPSCC have significantly better outcomes than patients with p16-negative disease. Although the result is intriguing and biologically plausible, p16 being an important tumor 4


suppressor for cellular function, the potential biologic or clinical significance of our results is unclear in this heterogeneous group of patients because of several limitations. Although the relevance of p16 expression as a prognostic biomarker is well established in the context of p16 as a surrogate for HPV infection, the significance of p16 decoupled from HPV is unclear. Numerous studies have developed and validated the scoring system used for defining p16 as positive (diffuse staining ⬎ 70% of tumor) in OPSCC, where HPV infection is common.5,6,9,10,23 However, use of this scoring system may not accurately reflect the significance of p16 expression independent of HPV infection in non-OPSCC. One commonly used method to develop an independent IHC scoring system is to assess an H factor (ie, intensity X percent positive tumor cells) of the p16 staining and identify an optimal threshold for separating the prognostic groups by using test and validation sets. In a recent study of oral cavity SCC, where 409 tumors were evaluated with multiple HPV detection methods, p16 expression was assessed with the H score, but again, the H score cutoff of 60 used was derived from an OPSCC study rather than independently derived from non-OPSCC.7,10 Unfortunately, our study is limited by an inability to determine H scores, because a majority of our data were generated using tissue microarrays, where the assessment of percent positive tumor cells in a representative core of tumor may not have accurately reflected the tumor overall. However, close evaluation of tumors with available wholetumor sections showed that only rare samples exhibited strong p16 staining in ⬍ 70% of the tumor cells (one [2%] of 50 tumors), and few JOURNAL OF CLINICAL ONCOLOGY


B

100

100

Overall Survival (%)

A

Progression-Free Survival (%)


80 60 40 20

p16 positive p16 negative

0

1

2

3

4

5

80 60 40 20

0

Time After Random Assignment (years) 62 260

45 167

38 131

34 106

23 74

No. at risk p16 positive p16 negative

10 23

D

100 80 60 40 20

1

2

3

4

62 260

5

19 178

16 142

15 116

100

10 80

F

40

20

1

2

3

No. at risk HPV positive HPV negative

4 30

60

0

4

16 2 22 142

13 2 18 110

12 2 16 90

7 2 12 63

57 212

44 170

41 140

28 91

13 33

2

3

4

5

40 20

5

28 269

HPV positive HPV negative

1

2 1 7 21

24 223

19 177

18 147

12 98

6 40

3

4

5

100

80

60

40

20

0

Time After Random Assignment (years) No. at risk HPV-pos p16 pos 20 HPV-pos p16 neg 7 HPV-neg p16 pos 33 HPV-neg p16 neg 213

5

Time After Random Assignment (years)

HPV-pos p16 pos HPV-pos p16 neg HPV-neg p16 pos HPV-neg p16 neg

80

4

60

0



E

28 269

3

80

Time After Random Assignment (years) No. at risk HPV positive HPV negative

2

100

HPV positive HPV negative

0

1



C




HPV-pos p16 pos HPV-pos p16 neg HPV-neg p16 pos HPV-neg p16 neg

1

2

Time After Random Assignment (years) No. at risk HPV-pos p16 pos 20 HPV-pos p16 neg 7 HPV-neg p16 pos 33 HPV-neg p16 neg 213

19 4 29 175

15 3 21 140

14 3 20 114

9 2 14 78

4 1 8 30

Fig 1. (A) Progression-free (PFS) and (B) overall survival (OS) by p16 expression for patients with nonoropharynx tumors. Patients with p16-positive tumors had significantly longer PFS (P ⫽ .04) and OS (P ⫽ .01) than patients with p16-negative tumors. Regarding PFS, 28 of 62 patients with p16-positive tumors and 160 of 260 patients with p16-negative tumors experienced progression; 5-year PFS estimates were 54.7% (95% CI, 41.6 to 67.8) and 34.3% (95% CI, 27.5 to 41.1), respectively. Regarding OS, 21 of 62 patients with p16-positive tumors and 134 of 260 patients with p16-negative tumors died; 5-year OS estimates were 64.4% (95% CI, 50.8 to 77.9) and 44.4% (95% CI, 37.4 to 51.4). (C) PFS and (D) OS by human papillomavirus (HPV) in situ hydridization (ISH) status for patients with nonoropharynx tumors. There was no significant difference in PFS (hazard ratio [HR], 0.77; 95% CI, 0.44 to 1.33; P ⫽ .35) or OS (HR, 0.64; 95% CI, 0.34 to 1.21; P ⫽ .17) between HPV ISH–positive and –negative patients. (E) PFS and (F) OS by HPV ISH status and p16 expression. There was no significant difference in PFS or OS (HPV positive/p16 positive v HPV positive/p16 negative v HPV negative/p16 positive v HPV negative/p16 negative: PFS, P ⫽ .23; OS, P ⫽ .21; HPV positive/p16 positive v HPV negative/p16 positive: PFS, P ⫽ .66; OS, P ⫽ .46; HPV positive/p16 positive v HPV negative/p16 negative: PFS, P ⫽ .20; OS, P ⫽ .09; HPV positive/p16 positive v HPV negative/p16 positive or p16 negative: PFS, P ⫽ .23; OS, P ⫽ .12).

www.jco.org



5

Chung et al

Table 3. Multivariable Analysis of PFS and OS End Point Variable PFS p16 status (positive v negative) Age (continuous) Sex (male v female) Clinical T stage (T4 v T1-3) Clinical N stage (N1-3 v N0/NX) OS p16 status (positive v negative) Age (continuous) Sex (male v female) Clinical T stage (T4 v T1-3) Clinical N stage (N1-3 v N0/NX)

HRⴱ

95% CI

P

0.63 1.02 1.44 1.38 2.63

0.42 to 0.95 1.00 to 1.03 1.00 to 2.07 1.01 to 1.88 1.78 to 3.89

.03 .03 .05 .04 ⬍ .001

0.56 1.03 1.62 1.55 2.18

0.35 to 0.89 1.01 to 1.05 1.07 to 2.45 1.11 to 2.17 1.44 to 3.28

.01 .002 .02 .01 ⬍ .001

Abbreviations: HR, hazard ratio; OS, overall survival; PFS, progression-free survival. ⴱ HRs estimated from Cox models stratified by protocol.

samples (three [6%] of 50 tumors) resulted in significant interobserver variability (positive v negative) between two pathologists (C.S.K., R.C.J.). This suggests that the OPSCC scoring system may also be applicable in non-OPSCC, but additional evaluation of the scoring system specific to non-OPSCC is required before broad clinical application. The rare cases of intermediate p16 staining (⬍ 70%) may be explained by data recently presented by the Cancer Genome Atlas (TCGA), which included a comprehensive analysis of 279 HNSCCs (HPV positive, n ⫽ 35; HPV negative, n ⫽ 244). In this extensive analysis, CDKN2A (gene coding p16 protein) was found to be inactivated in up to 90% of the HPV-negative tumors through mutation, deletion, abnormal splicing, or DNA methylation.24 This is consistent with historical data suggesting that the lack of p16 expression in HPV-negative tumors is not the result of undetectable expression of p16 protein, as in normal tissue,25,26 but rather the result of a lack of

functional CDKN2A and subsequent loss of p16 expression. The few tumors with intermediate p16 staining may be the result of incomplete loss of p16 expression or rare molecular alterations that require further investigation. The clinical significance of tumors with intermediate p16 staining is currently unknown, given the limited number of patient cases. Furthermore, determination of the prognostic value of p16 independent of HPV infection requires identification of true HPVnegative patients, but our study is hampered by limited tumor availability, which precluded HPV E6/7 RNA detection by the goldstandard qRT-PCR assay. This is a major limitation, because HPV ISH, used in our study, is known to be specific but less sensitive than other HPV detection methods.6,9,10,23 Our data show that 12% of non-OPSCCs were p16 positive/HPV negative, suggesting this group may represent a mixture of true HPV-positive patients with falsenegative HPV ISH results and true HPV-negative patients, in whom

Table 4. Interaction of p16 Status and Primary Site Model Oneⴱ End Point Comparison PFS p16 positive v negative Oropharynx Nonoropharynx Oropharynx v nonoropharynx p16 positive p16 negative P for interaction OS p16 positive v negative Oropharynx Nonoropharynx Oropharynx v nonoropharynx p16 positive p16 negative P for interaction

Model Two†

HR

95% CI

HR

95% CI

0.37 0.67

0.29 to 0.47 0.45 to 1.00

0.39 0.65

0.31 to 0.50 0.43 to 0.97

0.54 0.99

0.36 to 0.82 0.77 to 1.26

0.48 0.80

0.32 to 0.74 0.62 to 1.03

.01

.04

0.29 0.58

0.22 to 0.38 0.36 to 0.91

0.31 0.57

0.24 to 0.42 0.36 to 0.90

0.48 0.97

0.30 to 0.78 0.74 to 1.27

0.44 0.80

0.27 to 0.72 0.60 to 1.05

.01

.03

Abbreviations: HR, hazard ratio; OS, overall survival; PFS, progression-free survival. ⴱ Model one: HRs estimated from Cox models stratified by protocol, with covariates p16 status, primary site, and interaction of p16 status and primary site. †Model two: HRs estimated from Cox models stratified by protocol, with covariates p16 status, primary site, interaction of p16 status and primary site, age, sex, T stage, and N stage.

6





B 100

100



A 80

60

40 p16-pos op p16-pos nonop p16-neg op p16-neg nonop

20

0

1

2

3

4

5

80

60

40

20

0

Time After Random Assignment (years) No. at risk p16-pos op 493 p16-pos nonop 62 p16-neg op 198 p16-neg nonop 260

399 45 123 167

370 38 96 131

347 34 78 106

253 23 45 74

105 10 21 23

p16-pos op p16-pos nonop p16-neg op p16-neg nonop

1

2

3

4

5

Time After Random Assignment (years) No. at risk p16-pos op 493 p16-pos nonop 62 p16-neg op 198 p16-neg nonop 260

457 57 151 212

433 44 122 170

399 41 99 140

290 28 56 91

119 13 25 33

Fig 2. (A) Progression-free (PFS) and (B) overall survival (OS) by p16 expression and primary site (oropharyngeal squamous cell carcinoma [OPSCC] v non-OPSCC); 5-year PFS estimates were 69.3% (95% CI, 65.0% to 73.6%), 54.7% (95% CI, 41.6% to 67.8%), 36.8% (95% CI, 29.8% to 43.9%), and 34.3% (95% CI, 27.5% to 41.1%) for patients with p16-positive OPSCC (p16 pos op), p16-positive non-OPSCC (p16 pos nonop), p16-negative OPSCC (p16 neg op), and p16-negative non-OPSCC (p16 neg nonop), respectively; 5-year OS estimates were 79.8% (95% CI, 76.0% to 83.7%), 64.4% (95% CI, 50.8 to 77.9), 45.8% (95% CI, 38.1% to 53.5%), and 44.4% (95% CI, 37.4% to 51.4%) for patients with p16-positive OPSCC, p16-positive non-OPSCC, p16-negative OPSCC, and p16-negative non-OPSCC, respectively. There was significant interaction between p16 status and primary site for both PFS (P ⫽ .01) and OS (P ⫽ .01).

other molecular mechanisms led to overexpression of p16 in the absence of HPV infection. Therefore, p16 prognostic significance may be driven by HPV as well as RB1 loss, which can induce p16 overexpression. TCGA data show RB1 mutations in 4% of HPV-negative tumors.16,17,24 In addition, 2.6% of non-OPSCCs in our study were p16 negative/HPV positive, and TCGA data show that no HPVpositive tumors had CDKN2A inactivation.24 Even considering the relatively few HPV-positive tumors in the TCGA data set, our data suggest that the p16-negative/HPV-positive patient cases may be at least in part the result of false-negative p16. Therefore, additional research using multimodality testing in non-OPSCC and development of more accurate HPV testing are indicated. We have in addition observed differences in the concordance between p16 and HPV ISH among the subsites of the oral cavity, hypopharynx, and larynx, where oral cavity tumors have the worst concordance. This difference could be the result of the biologic and anatomic heterogeneity among these sites. For example, large tumors in the retromolar trigone may be indistinguishable from a tonsillar primary because of the anatomic proximity of these sites, resulting in misclassification. In addition, ectopic tonsillar tissue, albeit rare, has been reported in the floor of the mouth, where HPV infection can lead to the development of a tumor.27,28 Again, in the published study of oral cavity SCC, poor concordance between p16 and HPV ISH was primarily because of poor sensitivity of p16 IHC in detecting HPV.7 The PPV of HPV ISH was 95.2%, but the PPV of p16 IHC was only 40.4% when compared with HPV E6/7 RNA expression.7 Although this study determined that p16 is a poor surrogate marker of HPV infection in the oral cavity, consistent with our data, it did not evaluate whether p16 expression is prognostic independent of HPV status. In recent studies of oral tongue SCC (further delineating subsite even within oral cavity), Lim et al29 evaluated CDKN2A (p16) aberrations and did not observe a significant correlation with clinical characteriswww.jco.org

tics or outcome, whereas Harris et al30 reported a link between p16 expression and improved outcomes. These data demonstrate that each anatomic subsite needs separate evaluation with a larger sample size to clearly determine the prognostic role of p16 expression, and our findings should be interpreted with caution because of the small sample sizes in each site. In addition, presence of p16 expression may affect treatment outcomes and may have contributed to the survival differences seen in our study. In a recent study, p16 was shown to sensitize HPV-positive cells to ionizing radiation by inhibiting homologous recombination-mediated DNA damage response and downregulating cyclin D1 expression. The inhibition of the DNA damage response by p16 is thought to be independent of its cell-cycle regulation–inhibiting CDK4/6 activity.31 A majority of patients with oral cavity tumors in our study were treated with primary surgery followed by adjuvant therapy (68 [85%] of 80 in RTOG 0234), whereas a majority of patients with tumors in the larynx or hypopharynx were treated with primary concurrent chemoradiotherapy (161 [89%] of 181 and 54 [89%] of 61 in RTOG 0129 and 0522, respectively). Although we could not clearly delineate the prognostic impact of p16 expression in each subsite or in HPV-negative/p16-negative and HPVnegative/p16-positive groups because of the limited sample sizes, p16 as a biomarker independent of HPV status or potential biomarker of radiation sensitivity in non-OPSCC needs further evaluation. The clinically significant role of p16 expression as a surrogate marker of HPV infection for OPSCC is clearly established and appropriate. It is exemplified by its use as an integral biomarker for patient selection in clinical trials where there is a major effort to decrease the unnecessary toxicities of current treatment for HPV-positive patients without compromising survival. However, our data show that the use © 2014 by American Society of Clinical Oncology


7

Chung et al

of p16 IHC as a surrogate for HPV infection or as a prognostic biomarker in non-OPSCC requires further investigation before broad application in the clinical setting, and the role of p16 expression in radiation sensitivity warrants further investigation.

Bristol-Myers Squibb Research Funding: None Expert Testimony: None Patents, Royalties, and Licenses: None Other Remuneration: Maura L Gillison, Roche

AUTHOR CONTRIBUTIONS AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) and/or an author’s immediate family member(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a “U” are those for which no compensation was received; those relationships marked with a “C” were compensated. For a detailed description of the disclosure categories, or for more information about ASCO’s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: None Stock Ownership: None Honoraria: David Raben, Merck,

REFERENCES 1. Maier H, Dietz A, Gewelke U, et al: Tobacco and alcohol and the risk of head and neck cancer. Clin Investig 70:320-327, 1992 2. Gillison ML, D’Souza G, Westra W, et al: Distinct risk factor profiles for human papillomavirus type 16-positive and human papillomavirus type 16-negative head and neck cancers. J Natl Cancer Inst 100:407-420, 2008 3. Poling JS, Ma XJ, Bui S, et al: Human papillomavirus (HPV) status of non-tobacco related squamous cell carcinomas of the lateral tongue. Oral Oncol 50:306-310, 2014 4. Chaturvedi AK, Engels EA, Pfeiffer RM, et al: Human papillomavirus and rising oropharyngeal cancer incidence in the United States. J Clin Oncol 29:4294-4301, 2011 5. Ang KK, Harris J, Wheeler R, et al: Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 363:24-35, 2010 6. Rischin D, Young RJ, Fisher R, et al: Prognostic significance of p16INK4A and human papillomavirus in patients with oropharyngeal cancer treated on TROG 02.02 phase III trial. J Clin Oncol 28:41424148, 2010 7. Lingen MW, Xiao W, Schmitt A, et al: Low etiologic fraction for high-risk human papillomavirus in oral cavity squamous cell carcinomas. Oral Oncol 49:1-8, 2013 8. Lee SY, Cho NH, Choi EC, et al: Is human papillomavirus a causative factor of glottic cancer? J Voice 25:770-774, 2011 9. Shi W, Kato H, Perez-Ordonez B, et al: Comparative prognostic value of HPV16 E6 mRNA compared with in situ hybridization for human oropharyngeal squamous carcinoma. J Clin Oncol 27:6213-6221, 2009 10. Jordan RC, Lingen MW, Perez-Ordonez B, et al: Validation of methods for oropharyngeal cancer HPV status determination in US cooperative group trials. Am J Surg Pathol 36:945-954, 2012 11. Rietbergen MM, Brakenhoff RH, Bloemena E, et al: Human papillomavirus detection and comor-

8


Conception and design: Christine H. Chung, Richard C. Jordan, Quynh-Thu Le Financial support: Christine H. Chung Collection and assembly of data: Christine H. Chung, Qiang Zhang, Christina S. Kong, Jonathan Harris, Paul M. Harari, Dian Wang, Kevin P. Redmond, George Shenouda, Andy Trotti, David Raben, Maura L. Gillison, Richard C. Jordan, Quynh-Thu Le Data analysis and interpretation: Christine H. Chung, Qiang Zhang, Christina S. Kong, Jonathan Harris, Elana J. Fertig, Maura L. Gillison, Richard C. Jordan, Quynh-Thu Le Manuscript writing: All authors Final approval of manuscript: All authors

bidity: Critical issues in selection of patients with oropharyngeal cancer for treatment de-escalation trials. Ann Oncol 24:2740-2745, 2013 12. Posner MR, Lorch JH, Goloubeva O, et al: Survival and human papillomavirus in oropharynx cancer in TAX 324: A subset analysis from an international phase III trial. Ann Oncol 22:1071-1077, 2011 13. Serrano M, Hannon GJ, Beach D: A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4. Nature 366:704-707, 1993 14. Dyson N, Howley PM, Münger K, et al: The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 243:934-937, 1989 15. Münger K, Werness BA, Dyson N, et al: Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO J 8:4099-4105, 1989 16. Parry D, Bates S, Mann DJ, et al: Lack of cyclin D-Cdk complexes in Rb-negative cells correlates with high levels of p16INK4/MTS1 tumour suppressor gene product. EMBO J 14:503-511, 1995 17. Andl T, Kahn T, Pfuhl A, et al: Etiological involvement of oncogenic human papillomavirus in tonsillar squamous cell carcinomas lacking retinoblastoma cell cycle control. Cancer Res 58:5-13, 1998 18. Harari PM, Harris J, Kies MS, et al: Postoperative chemoradiotherapy and cetuximab for high-risk squamous cell carcinoma of the head and neck (RTOG 0234). J Clin Oncol 32:2486-2495, 2014 19. Ang KK, Zhang Q, Rosenthal DI, et al: A randomized phase III trial of concurrent accelerated radiation plus cisplatin with or without cetuximab for stage III to IV head and neck carcinoma: RTOG 0522. J Clin Oncol doi:10.1200/JCO.2013. 53.5633 20. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:467-481, 1958

21. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50:163-170, 1966 22. Cox DR: Regression model and life-tables. J R Stat Soc Series B 34:187-202, 1972 23. Thavaraj S, Stokes A, Guerra E, et al: Evaluation of human papillomavirus testing for squamous cell carcinoma of the tonsil in clinical practice. J Clin Pathol 64:308-312, 2011 24. Hayes DN, Grandis JR, El-Naggar AK: The Cancer Genome Atlas: Integrated analysis of genome alterations in squamous cell carcinoma of the head and neck. J Clin Oncol 31:365s, 2013 (suppl 15s; abstr 6009) 25. Montebugnoli L, Cervellati F, Cocchi R, et al: Immunohistochemical expression of p16(INK4A) protein as a helpful marker of a subset of potentially malignant oral epithelial lesions: Study on a series with long-term follow-up. Histopathology 57:528534, 2010 26. Queiroz AB, Focchi G, Dobo C, et al: Expression of p27, p21(WAF/Cip1), and p16(INK4a) in normal oral epithelium, oral squamous papilloma, and oral squamous cell carcinoma. Anticancer Res 30: 2799-2803, 2010 27. Patel K, Ariyaratnam S, Sloan P, et al: Oral tonsils (ectopic oral tonsillar tissue). Dent Update 31:291-292, 2004 28. Mogi K: Ectopic tonsillar tissue in the mucosa of the floor of the mouth simulating a benign tumour: Case report. Aust Dent J 36:456-458, 1991 29. Lim AM, Do H, Young RJ, et al: Differential mechanisms of CDKN2A (p16) alteration in oral tongue squamous cell carcinomas and correlation with patient outcome. Int J Cancer 135:887-895, 2014 30. Harris SL, Thorne LB, Seaman WT, et al: Association of p16(INK4a) overexpression with improved outcomes in young patients with squamous cell cancers of the oral tongue. Head Neck 33:16221627, 2011 31. Dok R, Kalev P, Van Limbergen EJ, et al: P16INK4a impairs homologous recombinationmediated DNA repair in human papillomaviruspositive head and neck tumors. Cancer Res 74: 1739-1751, 2014




Affiliations Christine H. Chung and Elana J. Fertig, Johns Hopkins University, Baltimore, MD; Qiang Zhang and Jonathan Harris, Radiation Therapy Oncology Group Statistical Center, Philadelphia, PA; Christina S. Kong and Quynh-Thu Le, Stanford University, Stanford; Richard C. Jordan, University of California at San Francisco, San Francisco, CA; Paul M. Harari, University of Wisconsin Carbone Cancer Center, Madison; Dian Wang, Medical College of Wisconsin, Milwaukee, WI; Kevin P. Redmond, University of Cincinnati, Cincinnati; Maura L. Gillison, Ohio State University, Columbus, OH; George Shenouda, McGill University Health Centre, Montreal, Quebec, Canada; Andy Trotti, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL; and David Raben, University of Colorado, Denver, CO. ■ ■ ■

GLOSSARY TERMS

immunohistochemistry: the application of antigenantibody interactions to histochemical techniques. Typically, a tissue section is mounted on a slide and incubated with antibodies (polyclonal or monoclonal) specific to the antigen (primary reaction). The antigen-antibody signal is then amplified using a second antibody conjugated to a complex of peroxidaseantiperoxidase, avidin-biotin-peroxidase, or avidin-biotin alkaline phosphatase. In the presence of substrate and chromogen, the enzyme forms a colored deposit at the sites of antibodyantigen binding. Immunofluorescence is an alternate approach to visualize antigens. In this technique, the primary antigenantibody signal is amplified using a second antibody conjugated to a fluorochrome. On ultraviolet light absorption, the fluorochrome emits its own light at a longer wavelength (fluorescence), thus allowing localization of antibody-antigen complexes. p16: molecule that binds to cyclin-dependent kinase 4 and 6, thereby preventing their interaction with cyclin D. p16 (also known as p16INK4) behaves as a negative regulator of proliferation and arrests cells in the G0/G1 phase of the cell cycle.

www.jco.org

polymerase chain reaction (PCR): a method that allows logarithmic amplification of short DNA sequences within a longer DNA molecule.

pRb (Rb phosphorylation): phosphorylated form of Rb, the retinoblastoma susceptibility gene product. Phosphorylated Rb has important ramifications for cell cycle progression. In the phosphorylated state, Rb is unable to bind to the transcriptional factor E2F (which is also important for cell cycle regulation). This results in an excess of free E2F, which can then induce transcription of genes involved in cell cycle progression. Hence, phosphorylation of Rb allows cells to progress through the G1 checkpoint into the S phase of the cell cycle. See Rb. quantitative real-time polymerase chain reaction (RT-PCR): process used in detecting polymerase chain reaction products as they accumulate that can be applied to gene expression quantification by reverse transcription of RNA into cDNA. In spite of its name (ie, quantitative), results are usually normalized to an endogenous reference. Devices allow the simultaneous assessment of many RNA sequences.



9

Chung et al

Appendix

Table A1. Characteristics of Patients With and Without p16 Expression Data RTOG 0129 p16 Unknown (n ⫽ 203) Characteristic Age, years Mean SD Median Minimum to maximum Q1 to Q3 Sex Male Female Zubrod performance status 0 1 Smoking history Never smoked Ever smoked Smoking history, pack-yearsⴱ Mean SD Median Minimum to maximum Q1 to Q3 Primary site Oral cavity Hypopharynx Larynx Clinical T stage T1 T2 T3 T4 TX Unknown Clinical N stage N0 N1 N2a N2b N2c N3 NX Unknown

No.

%

RTOG 0234

p16 Known (n ⫽ 85) No.

%

p16 Unknown (n ⫽ 34) No.

%

RTOG 0522

p16 Known (n ⫽ 95) No.

%


%

Total

p16 Known (n ⫽ 142) No.

%


%

p16 Known (n ⫽ 322) No.

%

56.1 8.46 57

57.1 9.71 56

57.0 10.84 58

56.0 10.25 57

57.0 8.30 56.5

58.1 8.59 59

56.5 8.64 57

57.2 9.41 58

26 to 82 51 to 61

30 to 81 51 to 63

32 to 76 49 to 63

27 to 77 50 to 63

34 to 77 51 to 62

31 to 79 52 to 65

26 to 82 51 to 62

27 to 81 51 to 64

170 33

83.7 16.3

61 24

71.8 28.2

25 9

73.5 26.5

67 28

70.5 29.5

101 23

81.5 18.5

116 26

81.7 18.3

296 65

82.0 18.0

244 78

75.8 24.2

106 52.2 97 47.8 n ⫽ 144 16 11.1 128 88.9

41 48.2 44 51.8 n ⫽ 61 5 8.2 56 91.8

15 44.1 19 55.9 n ⫽ 31 8 25.8 23 74.2

42 44.2 53 55.8 n ⫽ 73 17 23.3 56 76.7

75 60.5 49 39.5 n ⫽ 96 4 4.2 92 95.8

76 53.5 66 46.5 n ⫽ 128 7 5.5 121 94.5

196 54.3 165 45.7 n ⫽ 271 28 10.3 243 89.7

159 49.4 163 50.6 n ⫽ 262 29 11.1 233 88.9

n ⫽ 144 40.0 31.66 37.5

n ⫽ 61 43.2 28.42 40

n ⫽ 31 24.6 27.44 17.5

n ⫽ 73 33.8 29.96 33

n ⫽ 96 39.7 30.86 34

n ⫽ 128 42.6 28.74 40

n ⫽ 271 38.1 31.20 35

n ⫽ 262 40.3 29.18 40

0 to 137.5 14.925 to 58

0 to 111 24 to 56.25

0 to 114 0 to 42

0 to 120 1.5 to 51

0 to 162 20 to 52.5

0 to 135 20 to 60

0 to 162 15 to 53.75

0 to 135 17.6 to 60

30 38 135

14.8 18.7 66.5

12 20 53

14.1 23.5 62.4

27 5 2

79.4 14.7 5.9

68 7 20

71.6 7.4 21.1

0 28 96

0.0 22.6 77.4

0 34 108

0.0 23.9 76.1

57 71 233

15.8 19.7 64.5

80 61 181

24.8 18.9 56.2

0 29 115 59 0 0

0.0 14.3 56.7 29.1 0.0 0.0

0 10 46 29 0 0

0.0 11.8 54.1 34.1 0.0 0.0

4 9 7 10 2 2

11.8 26.5 20.6 29.4 5.9 5.9

12 28 23 32 0 0

12.6 29.5 24.2 33.7 0.0 0.0

0 23 79 22 0 0

0.0 18.5 63.7 17.7 0.0 0.0

0 35 76 31 0 0

0.0 24.6 53.5 21.8 0.0 0.0

4 61 201 91 2 2

1.1 16.9 55.7 25.2 0.6 0.6

12 73 145 92 0 0

3.7 22.7 45.0 28.6 0.0 0.0

69 31 12 35 42 14 0 0

34.0 15.3 5.9 17.2 20.7 6.9 0.0 0.0

30 20 2 15 16 2 0 0

35.3 23.5 2.4 17.6 18.8 2.4 0.0 0.0

7 2 3 11 8 1 0 2

20.6 5.9 8.8 32.4 23.5 2.9 0.0 5.9

15 31 7 20 14 2 2 4

15.8 32.6 7.4 21.1 14.7 2.1 2.1 4.2

34 14 2 32 40 2 0 0

27.4 11.3 1.6 25.8 32.3 1.6 0.0 0.0

35 21 9 23 51 3 0 0

24.6 14.8 6.3 16.2 35.9 2.1 0.0 0.0

110 47 17 78 90 17 0 2

30.5 13.0 4.7 21.6 24.9 4.7 0.0 0.6

80 72 18 58 81 7 2 4

24.8 22.4 5.6 18.0 25.2 2.2 0.6 1.2

Abbreviations: Q1, first quartile; Q3, third quartile; RTOG, Radiation Therapy Oncology Group; SD, standard deviation. ⴱ Pack-year is defined as equivalent of smoking one pack of cigarettes per day for 1 year.

10





Table A2. Patient Characteristics by p16 Expression RTOG 0129 p16 Negative (n ⫽ 73) Characteristic Age, years (P ⫽ .38)ⴱ Mean SD Median Minimum to maximum Q1 to Q3 Sex (P ⫽ .87)† Male Female Zubrod performance status (P ⫽ .02)† 0 1 Smoking history (P ⫽ .63)† Never smoked Ever smoked Smoking history: pack-years (P ⫽ .12)ⴱ‡ Mean SD Median Minimum to maximum Q1 to Q3 Primary site (P ⫽ .19)§ Oral cavity Hypopharynx Larynx Clinical T stage (P ⫽ .60)ⴱ T1 T2 T3 T4 Clinical N stage (P ⫽ .59)ⴱ N0 N1 N2a N2b N2c N3 NX Unknown

No.

%

57.2 9.86 55 30 to 81 51 to 63 52 21

RTOG 0234

p16 Positive (n ⫽ 12) No.

%

56.4 9.13 59.5 34 to 66 51 to 62.5 75.0 25.0

p16 Negative (n ⫽ 72) No.

%

56.3 10.70 58 27 to 77 50 to 63 52 20

72.2 27.8

RTOG 0522


%

55.0 8.85 55 36 to 69 49 to 63 15 8

65.2 34.8


%

58.4 8.48 59 31 to 79 52 to 65 92 23

80.0 20.0

Total


%

56.8 9.08 58 42 to 71 47 to 63

71.2 28.8

9 3

24 3

88.9 11.1

33 45.2 40 54.8 n ⫽ 51 4 7.8 47 92.2

8 4

66.7 33.3 n ⫽ 10 1 10.0 9 90.0

31 43.1 41 56.9 n ⫽ 55 14 25.5 41 74.5

11 47.8 12 52.2 n ⫽ 18 3 16.7 15 83.3

56 48.7 59 51.3 n ⫽ 101 4 4.0 97 96.0

20 74.1 7 25.9 n ⫽ 27 3 11.1 24 88.9

n ⫽ 51 45.9 29.51 40 0 to 111 25 to 70

n ⫽ 10 29.7 17.39 31.875 0 to 52.5 18 to 41

n ⫽ 55 31.9 28.15 33 0 to 110 0 to 50

n ⫽ 18 39.9 35.08 33.85 0 to 120 4 to 67.5

n ⫽ 101 45.1 29.00 44 0 to 135 25 to 61.5

n ⫽ 27 33.1 26.13 35 0 to 90 8 to 50


%

57.5 9.53 58 27 to 81 52 to 64 196 64

120 140 n⫽ 22 185

75.4 24.6

46.2 53.8 207 10.6 89.4

n ⫽ 207 41.8 29.38 40 0 to 135 20 to 60


%

56.1 8.89 57 34 to 71 50 to 63 48 14

77.4 22.6

39 62.9 23 37.1 n ⫽ 55 7 12.7 48 87.3 n ⫽ 55 34.7 27.96 33.75 0 to 120 10 to 51

8 20 45

11.0 27.4 61.6

4 0 8

33.3 0.0 66.7

51 6 15

70.8 8.3 20.8

17 1 5

73.9 4.3 21.7

0 25 90

0.0 21.7 78.3

0 9 18

0.0 33.3 66.7

59 51 150

22.7 19.6 57.7

21 10 31

33.9 16.1 50.0

0 8 39 26

0.0 11.0 53.4 35.6

0 2 7 3

0.0 16.7 58.3 25.0

8 21 21 22

11.1 29.2 29.2 30.6

4 7 2 10

17.4 30.4 8.7 43.5

0 29 60 26

0.0 25.2 52.2 22.6

0 6 16 5

0.0 22.2 59.3 18.5

8 58 120 74

3.1 22.3 46.2 28.5

4 15 25 18

6.5 24.2 40.3 29.0

28 18 2 11 12 2 0 0

38.4 24.7 2.7 15.1 16.4 2.7 0.0 0.0

2 2 0 4 4 0 0 0

16.7 16.7 0.0 33.3 33.3 0.0 0.0 0.0

13 22 4 16 10 2 1 4

18.1 30.6 5.6 22.2 13.9 2.8 1.4 5.6

2 9 3 4 4 0 1 0

8.7 39.1 13.0 17.4 17.4 0.0 4.3 0.0

26 18 7 20 41 3 0 0

22.6 15.7 6.1 17.4 35.7 2.6 0.0 0.0

9 3 2 3 10 0 0 0

33.3 11.1 7.4 11.1 37.0 0.0 0.0 0.0

67 58 13 47 63 7 1 4

25.8 22.3 5.0 18.1 24.2 2.7 0.4 1.5

13 14 5 11 18 0 1 0

21.0 22.6 8.1 17.7 29.0 0.0 1.6 0.0

Abbreviations: Q1, first quartile; Q3, third quartile; RTOG, Radiation Therapy Oncology Group; SD, standard deviation. ⴱ Wilcoxon rank-sum test comparing p16-negative and p16-positive total columns; NX and unknown were excluded. †Fisher’s exact test comparing p16-negative and p16-positive total columns. ‡Pack-year is defined as equivalent of smoking one pack of cigarettes per day for 1 year. §Pearson’s ␹2 test comparing p16-negative and p16-positive total columns.

www.jco.org



11

Chung et al

Table A3. Patient Characteristics by High-Risk HPV ISH Status RTOG 0129 HPV Negative (n ⫽ 87) Characteristic Age, years (P ⫽ .82)ⴱ Mean SD Median Minimum to maximum Q1 to Q3 Sex (P ⫽ .25)† Male Female Zubrod performance status (P ⫽ .33)† 0 1 Smoking history (P ⫽ .75)† Never smoked Ever smoked Smoking history: pack-years (P ⫽ .07)ⴱ‡ Mean SD Median Minimum to maximum Q1 to Q3 Primary site (P ⫽ .11)§ Oral cavity Hypopharynx Larynx Clinical T stage (P ⫽ .96)ⴱ T1 T2 T3 T4 Clinical N stage (P ⫽ .04)ⴱ N0 N1 N2a N2b N2c N3 NX Unknown

No.

%

57.3 9.19 57 30 to 81 51 to 62 62 25

RTOG 0234

HPV Positive (n ⫽ 6) No.

%

56.2 12.19 60.5 34 to 66 51 to 65 100.0 0.0

HPV Negative (n ⫽ 88) No.

%

55.4 10.68 57 27 to 77 49.5 to 63

71.3 28.7

6 0

60 28

68.2 31.8

43 49.4 44 50.6 n ⫽ 61 5 8.2 56 91.8 n ⫽ 61 44.6 29.31 40 0 to 111 25 to 60

4 2

66.7 33.3 n⫽5 0 0.0 5 100.0 n⫽5 25.6 16.66 18 10 to 52.5 17.5 to 30

38 43.2 50 56.8 n ⫽ 69 18 26.1 51 73.9 n ⫽ 69 32.7 30.59 32 0 to 120 0 to 50

12 20 55

13.8 23.0 63.2

2 0 4

33.3 0.0 66.7

64 6 18

0 9 47 31

0.0 10.3 54.0 35.6

0 2 1 3

0.0 33.3 16.7 50.0

33 20 2 14 16 2 0 0

37.9 23.0 2.3 16.1 18.4 2.3 0.0 0.0

0 2 0 0 3 1 0 0

0.0 33.3 0.0 0.0 50.0 16.7 0.0 0.0

RTOG 0522


%

56.7 9.79 59 30 to 67 49 to 65 12 3

80.0 20.0


%

59.1 8.29 60 36 to 79 52 to 65 78 16

Total


%

55.0 6.53 57 42 to 62 52 to 59 85.7 14.3


%

57.3 9.51 58 27 to 81 51 to 63 200 69

74.3 25.7


%

56.2 9.34 58.5 30 to 67 51.5 to 63

83.0 17.0

6 1

24 4

85.7 14.3

6 40.0 9 60.0 n ⫽ 12 1 8.3 11 91.7 n ⫽ 12 36.8 24.64 37.85 0 to 78 18.75 to 55

45 47.9 49 52.1 n ⫽ 88 4 4.5 84 95.5 n ⫽ 88 44.4 27.57 42 0 to 135 27.15 to 60

6 1

85.7 14.3 n⫽7 1 14.3 6 85.7 n⫽7 16.7 16.18 15 0 to 45 1.35 to 30

126 46.8 143 53.2 n ⫽ 218 27 12.4 191 87.6 n ⫽ 218 40.8 29.42 40 0 to 135 18 to 58

16 57.1 12 42.9 n ⫽ 24 2 8.3 22 91.7 n ⫽ 24 28.6 22.07 22 0 to 78 12.5 to 45

72.7 6.8 20.5

11 1 3

73.3 6.7 20.0

0 28 66

0.0 29.8 70.2

0 2 5

0.0 28.6 71.4

76 54 139

28.3 20.1 51.7

13 3 12

46.4 10.7 42.9

12 24 25 27

13.6 27.3 28.4 30.7

1 6 1 7

6.7 40.0 6.7 46.7

0 26 49 19

0.0 27.7 52.1 20.2

0 2 4 1

0.0 28.6 57.1 14.3

12 59 121 77

4.5 21.9 45.0 28.6

1 10 6 11

3.6 35.7 21.4 39.3

17 27 5 20 11 2 2 4

19.3 30.7 5.7 22.7 12.5 2.3 2.3 4.5

0 6 2 2 5 0 0 0

0.0 40.0 13.3 13.3 33.3 0.0 0.0 0.0

20 14 8 20 32 0 0 0

21.3 14.9 8.5 21.3 34.0 0.0 0.0 0.0

2 1 1 0 3 0 0 0

28.6 14.3 14.3 0.0 42.9 0.0 0.0 0.0

70 61 15 54 59 4 2 4

26.0 22.7 5.6 20.1 21.9 1.5 0.7 1.5

2 9 3 2 11 1 0 0

7.1 32.1 10.7 7.1 39.3 3.6 0.0 0.0

Abbreviations: ISH, in situ hybridization; Q1, first quartile; Q3, third quartile; RTOG, Radiation Therapy Oncology Group; SD, standard deviation. ⴱ Wilcoxon rank-sum test comparing p16-negative and p16-positive total columns; NX and unknown were excluded. †Fisher’s exact test comparing p16-negative and p16-positive total columns. ‡Pack-year is defined as equivalent of smoking one pack of cigarettes per day for 1 year. §Pearson’s ␹2 test comparing p16-negative and p16-positive total columns.

12





Table A4. Anatomic Subsites Included As Nonoropharynx Squamous Cell Carcinoma Subsite Oral cavity Oral tongue Floor of mouth Buccal mucosa Lower gingiva Upper gingiva Retromolar trigone Oral cavity, NOS Larynx Ventricular band Arytenoid Suprahyoid epiglottis Infrahyoid epiglottis Aryepiglottic fold Vocal cords Subglottis Supraglottic larynx, NOS Glottic larynx, NOS Subglottic larynx, NOS Hypopharynx Pyriform fossa Postcricoid area Posterior wall Hypopharynx, NOS

No. of Patients

%

40 19 6 3 3 6 3 3 3 7 7 6 12 20 4 98 23 1

12.4 5.9 1.9 0.9 0.9 1.9 0.9 0.9 0.9 2.2 2.2 1.9 3.7 6.2 1.2 30.4 7.1 0.3

34 2 7 18

10.6 0.6 2.2 5.6

Abbreviations: NOS, not otherwise specified.

www.jco.org



13

Chung et al

100


B Progression-Free Survival (%)

A HR (95% CI) [p16 positive/p16 negative] 0.80 (0.42 to 1.55) 2-sided log-rank P = .52

80

60

40

20 p16 positive p16 negative

0

1

2

3

4

100

HR (95% CI) [p16 positive/p16 negative] 0.33 (0.12 to 0.96) 2-sided log-rank P = .03

80

60

40


0

5

1

21 59

14 35

12 27

10 26

8 21

3

4

5


Time After Random Assignment (years) No. at risk p16 positive p16 negative

2

3 7

10 51

8 30

6 23

6 16

4 12

1 3

100


80

60

40


0

1

2

3

4


D Progression-Free Survival (%)

C


5

100

80

60

40


0


31 150

23 102

20 81

18 64

11 41

6 13


1

2

3

4

5


62 260

45 167

38 131

34 106

23 74

10 23

Fig A1. Progression-free survival by p16 expression for patients with (A) oral cavity, (B) hypopharynx, and (C) larynx tumors and (D) for these patients combined. HR, hazard ratio.

14





A

B 100




100

80

60

40

20


80

60

40



0

1

2

3

4

0

5

1

21 59

19 44

13 32

13 29

10 23

3

4

5



2

5 8

C


10 51

9 39

7 31

6 23

4 14

1 4

4

5

D 100



100

80

60

40


20

80

60

40

20


0

1


2

3

4

5

0


31 150

29 129


24 107

22 88

14 54

7 21

1

2

3


62 260

57 212

44 170

41 140

28 91

13 33

Fig A2. Overall survival by p16 expression for patients with (A) oral cavity, (B) hypopharynx, and (C) larynx tumors and (D) for these patients combined. HR, hazard ratio.

www.jco.org



15

Human papillomavirus detection in head and neck squamous cell carcinoma.

Human papillomavirus-related head and neck squamous cell carcinoma variants.

Frequency and prognostic significance of p16(INK4A) protein overexpression and transcriptionally active human papillomavirus infection in laryngeal squamous cell carcinoma.

A comprehensive evaluation of human papillomavirus positive status and p16INK4a overexpression as a prognostic biomarker in head and neck squamous cell carcinoma.

Human papillomavirus DNA and p16 expression in Japanese patients with oropharyngeal squamous cell carcinoma.

Elevated intrinsic cancer stem cell population in human papillomavirus-associated head and neck squamous cell carcinoma.

Head and neck squamous cell carcinoma and human papillomavirus: epidemiology, treatment and future trends.

Human papillomavirus and p53 mutations in head and neck squamous cell carcinoma among Japanese population.

Epidemiology of Human Papillomavirus-Positive Head and Neck Squamous Cell Carcinoma.

Squamous cell carcinoma of the head and neck outside the oropharynx is rarely human papillomavirus related.

Low prevalence of human papillomavirus in head and neck squamous cell carcinoma in Chinese patients.

Expression of cdk6 in head and neck squamous cell carcinoma.

Expression profile and in vitro blockade of programmed death-1 in human papillomavirus-negative head and neck squamous cell carcinoma.

Survivin and pAkt as potential prognostic markers in squamous cell carcinoma of the head and neck.

Comorbidity and performance status as independent prognostic factors in patients with head and neck squamous cell carcinoma.

Circulating tumor cells as biomarkers of head and neck squamous cell carcinoma: an updated view.

Fibroblast Growth Factor Receptor Family Members as Prognostic Biomarkers in Head and Neck Squamous Cell Carcinoma: A Systematic Review.

The prognostic value of cyclin D1 expression in head and neck squamous cell carcinoma.

Circulating microRNAs as novel prognosis biomarkers for head and neck squamous cell carcinoma.

Prognostic significance of p16 and its relationship with human papillomavirus in pharyngeal squamous cell carcinomas.

IMP3 and p16 expression in squamous cell carcinoma of the head and neck: A comparative immunohistochemical analysis.

Expression and role of the embryonic protein SOX2 in head and neck squamous cell carcinoma.

Rb pathway.

The prognostic value of Tiam1 protein expression in head and neck squamous cell carcinoma: a retrospective study.